Return to site

The Implications of Machine Learning on Condensed Matter Physics & Quantum Computing

· Machine Learning,Condensed Matter,Quantum Computing,Physics,Ai

The Implications of Machine Learning on Condensed Matter Physics & Quantum Computing

With all the advancements in machine learning in industry and medicine, it isn’t surprising that machines are also making their mark on the world of physics. More specifically, machine learning is aiding in the understanding and progression of the field that has the potential to give us super powerful batteries and efficient methods of power transmission [1]. This field is known as condensed-matter physics.

Condensed-matter physics is the field of physics that deals with the different properties of matter, with a focus on condensed phases under varying conditions.

Scientists in this field seek to understand the behavior of these different phases using the different laws of physics, including quantum mechanics. Quantum mechanics offers the complex concept of the wave function, which describes the state of a system as well as the probability of observing a particle at any given point, which would then provide insight as to the behavior of the system.

However, actually finding the wave function of a system, specifically a many-particle system, is one of the most complex problems physicists face today. To help solve this issue, they are beginning to turn to machine learning that opens the door to new innovation.

Juan Carrasquilla, a condensed matter theorist, is one of the few to have experimented with machine learning to determine the wave function of a system. He recognized that there are similarities between the algorithms used in condensed-matter physics and those used in AI.

The most common similarities include the way one could use an assumption for a function or problem of interest, then tune the parameters defining that assumption in order to get the algorithm to solve the problem.

Creating an algorithm that can find the wave function, Carrasquilla believes, could lead to solutions in the realms of materials science and chemistry that could, in turn, lead to significant discoveries in various different fields of research [2].

What does this mean for condensed-matter physics? Well, an understanding of the wave function with sufficient enough accuracy could lead to the development of new quantum materials and devices, and eventually, the first true quantum computer.

These computers have the potential to outperform the supercomputers we have today and will spur huge technological advancements in the fields of security, pharmaceuticals, and finance.

Such a computer is yet to exist, but IBM has created a computer in their quantum lab that comes pretty close. A perfect quantum computer is just a concept right now, but they are hopeful that even an imperfect quantum computer is still a useful one [3].

While it is still unsure what would be done with the first perfect quantum computer, there is hope and potential for application in financial markets, biology, and even furthering the advancement of artificial intelligence.

What is certain, however, is that the advancement of machine learning and application to condensed-matter physics is capable of bringing the world closer to perfect quantum computer and all of the technological advancements that come with it.

Written by Rachel Weissman & Edited by Alexander Fleiss

[1] Melko, R. (2017, February 01). The most complex problem in physics could be solved by machines with brains. Retrieved from https://qz.com/897033/applying-machine-learning-to-physics-could-be-the-way-to-build-the-first-quantum-computer/

[2] Q&A: A Condensed Matter Theorist Embraces AI. (2018, February 09). Retrieved from https://physics.aps.org/articles/v11/15

[3] Knight, W. (2018, March 04). Serious quantum computers are finally here. What are we going to do with them? Retrieved from https://www.technologyreview.com/s/610250/hello-quantum-world/